lq-preview · First MLX build 13 Oct 2026
LFM2.5-24B-A2B-Preview
MLX
Text lfm1.0
A 24B parameter text generation model from lq-preview, with 8 MLX builds, from 4-bit to bf16.
- Parameters
- 24B From Hugging Face
- Quantisation
- bf16 LFM2.5-24B-A2B-Preview-bf16
- Memory needed
- 56 GB Estimate, unified memory
- Download
- 47.9 GB Weights, estimated
- Downloads
- 473 This build, last 30 days
- Task
- Text generation Text
- MLX builds
- 8 On mlx-community
- Devices
- Mac Apple silicon
Memory
Will it fit?Needs about 56 GB at bf16
8 GB
16 GB
24 GB
32 GB
64 GB
128 GB
Fits with headroomTight, close other appsWon’t fit
Estimated from the size of the weights, plus a fifth for context and working memory and 2 GB for macOS and your apps. Long contexts need more.
Run it
Terminal · mlx-lm$ pip install -U mlx-lm
$ mlx_lm.generate --model mlx-community/LFM2.5-24B-A2B-Preview-bf16 --prompt "Hello"MLX builds
8 on mlx-community| LFM2.5-24B-A2B-Preview-4bit | 4-bit | 13.5 GB | 18 GB | 359 |
|---|---|---|---|---|
| LFM2.5-24B-A2B-Preview-mxfp4 | MXFP4 | 13.5 GB | 18 GB | 281 |
| LFM2.5-24B-A2B-Preview-nvfp4 | 4-bit | 13.5 GB | 18 GB | 161 |
| LFM2.5-24B-A2B-Preview-5bit | 5-bit | 16.5 GB | 21 GB | 173 |
| LFM2.5-24B-A2B-Preview-6bit | 6-bit | 19.5 GB | 24 GB | 163 |
| LFM2.5-24B-A2B-Preview-8bit | 8-bit | 25.5 GB | 31 GB | 168 |
| LFM2.5-24B-A2B-Preview-mxfp8 | MXFP8 | 25.5 GB | 31 GB | 145 |
| LFM2.5-24B-A2B-Preview-bf16 | bf16 | 47.9 GB | 56 GB | 473 |
Sizes are estimated from Hugging Face’s parameter counts and the quantisation. Downloads are for the last 30 days.